Change Details

This patch introduces an abstraction layer for TableGen backends. It separates the code generation logic and the syntax string output into separated modules. One module is the backend (`RegisterInfo` in this patch), the other an implementation of the abstract class `PrinterInterface`. The code generation stays in the TableGen backend. The generated syntax string however is written by an implementation of `PrinterInterface` to the output stream. Whenever the backend has generated a piece of code it requests the `PrinterInterface` to write the syntax to the output stream. The `PrinterInterface` only gets the necessary objects from the backend (e.g. Records to write, Register names etc.) and writes the syntax string to the output stream. Before this patch the TableGen backend wrote the generated code and its syntax on its own to the output stream. This had the disadvantage that no other languages, then the default C++ syntax, could be output easily. With the separation of the code generation (backend) and syntax output (`PrinterInterface`) it is easier to output syntax of other languages. It is only necessary to implement the methods of the `PrinterInterface` class and emit the language syntax which is needed. In this patch the `RegisterInfo` backend was refactored to use the `PrinterInterface` (`PrinterInterface` implementation in `PrinterLLVM.cpp`) as output module. If this gets upstreamed I will refactor other backends over this and the next month. For more background info you can take a look at: _Feedback implementation_ https://reviews.llvm.org/D136808 _Discussion in discourse.llvm.org_ https://discourse.llvm.org/t/comments-needed-for-refactoring-decoderemitter-tablegen-backend/65738

This patch introduces an abstraction layer for TableGen backends. It separates the code generation logic and the syntax string output into separated modules. One module is the backend (`RegisterInfo` in this patch), the other `PrinterLLVM`. The code generation stays in the TableGen backend. The syntax generation however is moved to `PrinterLLVM`, which writes the result to the output stream. Whenever the backend has generated a piece of code it requests `PrinterLLVM` to write the syntax to the output stream. Before this patch the TableGen backend generated the syntax on its own. This had the disadvantage that no other languages, then the default C++ syntax, could be output easily. With the separation of the code generation (backend) and syntax generation (`PrinterLLVM`) it is easier to output syntax of other languages. It is also possible write a printer which emits only specific parts of the generated code. A new printer only implements the methods of the `PrinterLLVM` it needs and emits its language syntax. In this patch the `RegisterInfo` backend was refactored to use `PrinterLLVM` as output module. If this gets upstreamed I can provide the refactor for the following modules in the next days: - DecoderEmitter - SubtargetInfo - InstructionInfo - AsmMatcher For more background info and where this idea is coming from you can take a look at: _Feedback implementation_ https://reviews.llvm.org/D136808 _Discussion in discourse.llvm.org_ https://discourse.llvm.org/t/comments-needed-for-refactoring-decoderemitter-tablegen-backend/65738

This patch introduces an abstraction layer for TableGen backends. It separates the code generation logic and the syntax string output into separated modules. One module is the backend (`RegisterInfo` in this patch), the other an implementation of the abstract class `PrinterInterface`LLVM`. The code generation stays in the TableGen backend. The generated syntax stringsyntax generation however is written by an implementation of `PrinterInterface`moved to `PrinterLLVM`, which writes the result to the output stream. Whenever the backend has generated a piece of code it requests the `PrinterInterface` to write the syntax to the output stream. The `PrinterInterface` only gets the necessary objects from the backend (e.g. Records to write, Register names etc.) and writes the syntax string`PrinterLLVM` to write the syntax to the output stream. Before this patch the TableGen backend wrote the generated code and itsthe syntax on its own to the output stream. This had the disadvantage that no other languages, then the default C++ syntax, could be output easily. With the separation of the code generation (backend) and syntax outputgeneration (`PrinterInterface`LLVM`) it is easier to output syntax of other languages. It is only necessary toalso possible write a printer which emits only specific parts of the generated code. A new printer only implements the methods of the `PrinterInterface` clasLLVM` it needs and emit thes its language syntax which is needed. In this patch the `RegisterInfo` backend was refactored to use the `PrinterInterface` (`PrinterInterface` implementation in `PrinterLLVM.cpp`) as output module.`PrinterLLVM` as output module. If this gets upstreamed I can provide the refactor for the following modules in the next days: - DecoderEmitter - SubtargetInfo - InstructionInfo If this gets upstreamed I will refactor other backends over this and the next month.- AsmMatcher For more background info and where this idea is coming from you can take a look at: _Feedback implementation_ https://reviews.llvm.org/D136808 _Discussion in discourse.llvm.org_ https://discourse.llvm.org/t/comments-needed-for-refactoring-decoderemitter-tablegen-backend/65738